Practical Multivariate Data Correction and Sensor Calibration Methods

We describe some practical and empirically validated multivariate sensor calibration methods, which have allowed us to successfully obtain usable data from low-cost, low-precision sensors and signal conditioning electronics; we demonstrate the efficacy of these methods with real examples of nonlinear correction of thermocouple measurements from errors originated by the simultaneous influence of the cold junction and the temperature of the signal conditioning and digitizing electronic circuit components.